@polyweatheryuan It's probably from WU's internal infrastructure and some nasty caching or data pipeline normalisation (etc). It's _eventually_ consistent but I agree is a bit of a problem for resolution (in a timely manner). Looks to be no insider trading which is nice though. Difficult problem.
@alejoamiras@aztecnetwork ZK tax would be awesome but I doubt governments would accept not knowing (and snooping) without a mass societal shift in how we think about privacy too. This one specifically feels like one of the more complicated issues.
Congratulations to the Obsidion team, particularly @michaelelliot and @madztheo who I've had the pleasure of meeting on more than one occasion, for recently joining @AztecLabs_ . Very proud of what you've achieved, if not a little envious too >:)
@zkpedrongmi@NoirLang I wish I had the opportunity to be as closely involved as you were at least but it was still fun being a fly on the wall to a fly on the wall, despite no direct contribution to RSA and BigInt.
@savio_sou Comparisons referring to a scaled single-size input for comparison to the equivalent total input size of iterations for a scenario.
e.g. 64 bytes iterated 100 times would have a comparison scenario of a single input sized at 6,400 bytes.
Did someone say benchmarks (@savio_sou)?
Here's some Noir/bb benchmarks for hash functions (Poseidon2, Keccak256, SHA256) at input sizes ranging from 1 to 8192 bytes, iterated on 1, 10, and 100 times, and with a fixed single-sized input for comparison!
https://t.co/JGQ2cFu94p
@savio_sou Iterations referring to the number of distinct, touched (used) inputs to the circuit. These must be distinct and used to prevent the Noir compiler optimising their use away when compiling the circuit for execution.
e.g. 32 bytes iterated 10 times (10 inputs of 32 bytes).